Efficient electric boat

Hi Jeremy,

Instead of using chain, have you considered using sprocket belt ? Lots of advantages and it should be slightly more efficient than chain.

Also keep engineering plastics in mind, some of them are really low resistant an slow to wear if any.

Question - Once you think you have achieved the optimal setup, what then
Would be nice if you would in the long run take it even further, for instance magnets of course offer the possibility of zero friction bearings

 

I've experimented with HTD toothed belts and used one as a primary reduction, but there's no doubt that the chain is much more efficient, perhaps half the power loss of even a small, 3mm pitch HTD, and around a quarter of the power loss of a 5mm HTD belt.

My experience is supported by the manufacturer's data, which does point out that these belts can be *almost* as efficient as a well-adjusted and lubricated chain. The main reason for using chain in the leg design was that it would be near-impossible to build and assemble with a belt, as the chain needs to be threaded around the sprockets and then connected. Another good reason for choosing chain is it's very high torque rating, much higher than a toothed belt of substantially greater dimensions.

The belt drive does have some significant advantages, including relative freedom from maintenance, no need for lubrication etc, and I can tolerate the small power loss it gives on the primary drive.

Bearing losses are tiny, too small to worry about investing time in reducing. Seal losses are significant, around 3 watts is lost on the main output shaft seal. The major sources of power loss are the prop efficiency (maybe 80% to 85% at best) and the motor itself (around 90%). The electrical losses will be around 5 to 10 watts, the belt drive is about 4 or 5 watts and the seal is another 3 watts. Bearing losses are immeasurably small on my test setup, so less than 1 watt. Total lost power for 70 watts delivered to the water should be around 35 to 40 watts, I think, so overall efficiency will be something like 65%, which is significantly better than the efficiency that a conventional ICE engine set up will give.

Jeremy

 

The highest efficiency motors back in the old days of my haunts on the power assist list were the ultra high rpms, and I always wondered why. Seems like remember a small, electronically commutated, PM rotor, model plane motor that ran at 80,000+! rpm with 92%? efficiency at the top of the list. I remember thinking that I wouldn't expect it because of bearing losses at that speed.

There was also a proposal for an electric "Wankel" motor that was supposed to be much more efficient in the larger, automobile power sizes. It was powered through about 80% of each rotation by an ever expanding PM rotor-PM stator gap releasing magnetic compression, similar to the compression cycle in the gasoline Wankel. At the top of the cycle a small electromagnet pulsed a reversed pole to initiate the following magnetic compression cycle. It did require a starter, flywheel, and ganged configuration for smoother power. Never amounted to anything apparently.

One theoretical thing I am curious about is what happens to the efficiency with superconducting wire. If you only gain another 9% or so, it hardly seems worth the trouble, at least on these smaller scales.

Guess I getting carried away with excitement on this discussion, excuse the dribble...

Porta

 

Interesting Jeremy, but if I add all those losses up the motor is going to spin freely in reverse

I recently heard of bearings made that can run in water. I got the impression thy were made from some engineering grade plastic in which case the water would act as a lubricant, but I could be wrong. Unfortunately I didn't persue the knowledge further.

If that is the case however, you could do away with seals.

As for the drive, you should give it a more hydrodynamic shape to least resistance in the water. The microlite guys have such a profile (a few sizes too) and it has a round 'tube' going through it as well. It may well be used for such an experiment.

Another thing that may prove an energy saver is if you can shape the drive like a wead eater pipe at the bottom. The motor and first shaft sits with an angle slanted aft, and the final drive and prop sits horizontal. Use a universal coupling to connect the two shafts. In a setup like this the submersable bearing may be an option.

 

Brilliant jeramy ..just ordered one from Hobby city ..we shall see ..thanks

 

I've been doing some looking at the RC hobby equipment offering for an auxiliary drive for a small sailboat. RC car controllers provide forward - reverse control. One supplier, Castle Creations, offer their controllers programmed for industrial use, including taking a throttle signal from a pot instead of the 'normal' servo signal. These are still modestly priced and might be more appropriate for our use. More info at http://www.castlespecialprojects.com/

 

I used a PTFE water lubricated bearing as the outboard drive shaft support bearing in the first test unit, the nice streamlined one pictured in this post: http://www.boatdesign.net/forums/boat-design/efficient-electric-boat-27996-2.html#post282904 earlier in this thread, so I fully understand the need to reduce drag. I just haven't got around to fitting the fairing to the chain drive, that's all.

PTFE bearings work well on slow speed shafts, in fact the shaft wears rather than the bearing. The frictional losses are low, too, for a plain bearing. Seals are needed to stop water getting into the working parts, like the gearbox, universal joint, chain or whatever, so are a necessary evil, despite their tendency to rob a bit of power.

The double universal joint option is being built now, I should have it ready for testing by the beginning of next week. So far it looks reasonably promising, at angles up to around 75 degrees or so the losses feel low in the simple test rig I made up to try it.

BTW, as it happens I have designed microlight and light aircraft, in the past..................

Jeremy

 

Nice find, Denny. It'd be good to get an "off the shelf" solution that doesn't need special modification. Castle Creations seem to have a very good reputation too, which bodes well.

Jeremy

 

Any of them fly ?... or is there a reason why you've stopped Just kiddin. We all have lots of designs and ideas that never fly. The wife says if I live to 500 I'll never get all my projects done. Won't mind tho, and not because there's more females now than 20 years ago

Post when you have it running, it would be interestig to see.

 

One design won "Outstanding Light Plane" at Sun'n'Fun (probably the World's second biggest aircraft show), and "Grand Champion" at the same show a year later, so I guess they do fly OK............

I managed to mix aircraft stuff and boating once, too, http://www.channel4.com/science/microsites/S/scrapheap2007/airboats_2.html

Jeremy

 

Sensorless

Hi Jeremy,

I'm following your very interesting project since some time and I'm really
impressed by your efforts.

I'm working on a similar project of a solar driven boat since two years,
although my progress is much slower. I will present my project here soon when I find more time.

Already a quick question now:
Did you already compare the sensorless and sensored mode on the big motor ?

I'm using a HXT63-74 motor that Ralph (powercroco) rewinded for me to be more efficient and slower but have problems in starting the motor in the water when using a sensorless RC-controller. I'm now starting to put Hall Sensors inside following your description in the endless sphere forum.

Best regards,

Thomas

 

Hi Thomas, thanks for the kind words.

I've run a couple of the TowerPro motors on my experimental drives, one fitted with Hall sensors, one without. Both start OK in water, although the Hall one will run much more slowly, with a massive amount of low speed torque compared to the sensorless one.

One thing that makes starting difficult when the motor is running without Hall sensors is if there is a high level of friction/stiction in the drive train. Some of the aircraft ESCs seem to struggle when starting with a load, although this shouldn't be a problem, even with a water prop, as the starting torque demand is very low if the bearings, seals, reduction drive etc are all able to run freely.

I've recently acquired a very similar motor to yours, a Turnigy 63-74 SK, re-wired from delta to star to lower the Kv to about 98. This seems to start OK using a cheap ESC, even with a modest load on the shaft.

I'm currently working on a direct drive unit, as the trade off between the reduction drive losses and the added I²R losses from running at a low voltage (12.8V), seem to work in favour of using direct drive. The current plan is to fit the motor in a sealed pod, with a nice streamlined shape (see photo below of the mould buck).

Jeremy

 

Hi Jeremy,

thanks for your quick reply.

In the moment I'm also trying a direct drive unit, although after the discussion I read here I was already planning to switch to a geared drive.

The attached photo shows the non-streamlined pod, in which I can fit the
63-74 motor (it replaces the original stator holder)

The original 63-74 motor was to fast (200kv) and therefore I tried to rewind it to get it slower. However than it failed to run, mainly as the Ri is getting to high and the sensorless driver does not get the right timing.

In the moment
the winding has around 160kv in delta mode and 100 in star mode, but does only work in the delta mode. This might be related to the fact that the flux ring is slightly distorted and therefore the distance to the magnets varies and/or that the high Ri gives trouble to the sensorless driver.
A new motor is ordered to replace the flux ring... but unfortunatly it is on backorder at hobbycity.com.

For the 160kv, the attached prop is much to big and I got currents > 70A at 12V... and therefore terrible losses.

Also attached are photos of the hull. It is modified after a Geodesic Airolite plan (http://www.gaboats.com/), can be divided into two parts and weights around 50lbs.

In the moment I'm working on very light solar panels (polycristalline cells, laminated between Teflon and EVA)... the batteries are 4x 50Ah LiFePo from HiPower and I just fitted the Fechter and Goodrum BMS.


I'm curious to hear how your different drive units compare to each other.

All the best,
Thomas

 

Nice engineering, Thomas. I love the look of that Airolite canoe.

My limited experience with the delta/star mod on a couple of standard motors has been that the ordinary RC controllers worked OK. The cheap sensorless controllers I have are 120 amp Swang Hobby ones. The sensored controller is a modified Infineon ebike controller, fitted with lower Rdson FETs (IRFB3077's). I've got a few sensorless adapter boards to test with the Infineon controller, to see if it will drive a standard RC outrunner OK without the need to fit sensors.

I'd be very interested to hear more about your solar cell work, it sounds very interesting. I would like to fit around 2m² of cells to the deck space I have on the boat I'm having built, and, because they are on the deck, would like to avoid the hazards of using glass.

Jeremy

 

Hi Jeremy,

Following your advice I also bought the smallest Infineon ebike controller to start with.

Some specific questions to get quicker started on the weekend (although they are slightly offtopic here):

- Can I just add a 100 Ohm potentiometer as throttle with the wiper connection connected to the green wire to supply 0-5V ?

- The programming connection on my controller (EB206) looks different than the one I saw in the endless spheres
forum. Do you have by chance the same controller and already know the pin layout of the programming connector ?

- The hall sensors (SS40A) have three connections, + - and 0... I guess that 0 are the signals and will connect
to the colored wires (yellow, green, blue)... Does the order/direction matter ?

Other question: Will you put some gearbox in your direct drice pod and which RPM do you aim for ?

Concerning the solar cells, I will make some photos and a description and add them here soon.

Thomas